LATER STAGES OF EVOLUTION OF IGNEOUS ROCKS 69 



gradually as they grew in the Kquid. The liquid passed through 

 all intermediate compositions toward the granitic and fed at various 

 stages the dykes and flows of the different t3^es of porph3n'y, 

 naturally later than the diabasic t5rpes. But in the visible upper 

 portion of the very large bodies of the magma, only those products 

 which were precipitated at the latest (and lightest), stage of the 

 magma's history are represented, i.e., the constituents of the rocks 

 of the granite family. 



Moreover, it seems quite reasonable and consistent to beheve 

 that the descent of igneous rocks has not been different from that 

 outhned above, even in the earliest stages of the earth's history of 

 which we have any knowledge from actual rock outcrops. The 

 rocks described by Sederholm belong among the oldest. In the 

 Canadian shield also the colossal bathoHths of the Laurentian were 

 preceded by the basaltic outpourings of the Keewatin, and the 

 same interpretation of their relation applies as that just given for 

 the Scandinavian shield. In the basement complex of Rhodesia 

 the contrasted volume of exposed acid and basic t3^es^ is attribut- 

 able to the action of similar processes. It seems possible, then, to 

 consider that basaltic magma is the original material of all igneous 

 action of which we have any direct knowledge, and that all igneous 

 rocks, even the huge areas of pre-Cambrian granitic rocks are 

 differentiates of basaltic magma. 



The objection might be raised that, in the more deeply eroded 

 terranes such as some pre-Cambrian areas must be, erosion should 

 have cut through the salic types and exposed their complementary 

 rocks if the salic bathohths are really only the upper portions of 

 bodies formed by differentiation of basaltic magma. In other 

 words, one would expect to find areas of basic rocks of batholithic 

 dimensions. The objection seems, however, to lose much of its 

 force when carefully considered. The granitic differentiate of an 

 originally basaltic magma may be as much as 10 or 15 per cent of 

 the total differentiated mass.^ Erosion would, therefore, have 

 to penetrate 10 or 15 per cent of the depth of the igneous mass to 



^ F. P. Mennel, A Manual of Petrology (London, 1913), p. 105. 

 ^ Collins has, for example, estimated the apUtic portions of the coarser Nipissing; 

 intrusives as 10 per cent or more of the total {Geol, Survey Canada, Mem. 33, 19 13, p. 65) ^ 



